the mtor effectors 4ebp1 and s6k2 are frequently...

The mTOR effectors 4EBP1 and S6K2 are

frequently coexpressed, and associated with a

poor prognosis and endocrine resistance in

breast cancer: a retrospective study including

patients from the randomised Stockholm

tamoxifen trials

Elin Karlsson, Gizeh Pérez-Tenorio, Risul Amin, Josefine Bostner, Lambert Skoog, Tommy

Fornander, Dennis C Sgroi, Bo Nordenskjöld, Anna-Lotta Hallbeck and Olle Stål

Linköping University Post Print

N.B.: When citing this work, cite the original article.

Original Publication:

Elin Karlsson, Gizeh Pérez-Tenorio, Risul Amin, Josefine Bostner, Lambert Skoog, Tommy

Fornander, Dennis C Sgroi, Bo Nordenskjöld, Anna-Lotta Hallbeck and Olle Stål, The mTOR

effectors 4EBP1 and S6K2 are frequently coexpressed, and associated with a poor prognosis

and endocrine resistance in breast cancer: a retrospective study including patients from the

randomised Stockholm tamoxifen trials, 2013, Breast Cancer Research, (15), 5, R96.

http://dx.doi.org/10.1186/bcr3557

Copyright: BioMed Central

http://www.biomedcentral.com/

Postprint available at: Linköping University Electronic Press

http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-104178

http://dx.doi.org/10.1186/bcr3557

http://www.biomedcentral.com/

http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-104178

http://twitter.com/?status=OA Article: The mTOR effectors 4EBP1 and S6K2 are frequently coexpressed, and associated wit... http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-104178 via @LiU_EPress %23LiU

Karlsson et al. Breast Cancer Research 2013, 15:R96http://breast-cancer-research.com/content/15/5/R96

RESEARCH ARTICLE Open Access

The mTOR effectors 4EBP1 and S6K2 arefrequently coexpressed, and associated with apoor prognosis and endocrine resistance inbreast cancer: a retrospective study includingpatients from the randomised Stockholmtamoxifen trialsElin Karlsson1*, Gizeh Pérez-Tenorio1, Risul Amin1, Josefine Bostner1, Lambert Skoog2, Tommy Fornander3,Dennis C Sgroi4, Bo Nordenskjöld1, Anna-Lotta Hallbeck1 and Olle Stål1

Abstract

Introduction: mTOR and its downstream effectors the 4E-binding protein 1 (4EBP1) and the p70 ribosomal S6kinases (S6K1 and S6K2) are frequently upregulated in breast cancer, and assumed to be driving forces intumourigenesis, in close connection with oestrogen receptor (ER) networks. Here, we investigated these factors asclinical markers in five different cohorts of breast cancer patients.

Methods: The prognostic significance of 4EBP1, S6K1 and S6K2 mRNA expression was assessed with real-time PCRin 93 tumours from the treatment randomised Stockholm trials, encompassing postmenopausal patients enrolledbetween 1976 and 1990. Three publicly available breast cancer cohorts were used to confirm the results.Furthermore, the predictive values of 4EBP1 and p4EBP1_S65 protein expression for both prognosis and endocrinetreatment benefit were assessed by immunohistochemical analysis of 912 node-negative breast cancers from theStockholm trials.

Results: S6K2 and 4EBP1 mRNA expression levels showed significant correlation and were associated with a pooroutcome in all cohorts investigated. 4EBP1 protein was confirmed as an independent prognostic factor, especiallyin progesterone receptor (PgR)-expressing cancers. 4EBP1 protein expression was also associated with a poorresponse to endocrine treatment in the ER/PgR positive group. Cross-talk to genomic as well as non-genomicER/PgR signalling may be involved and the results further support a combination of ER and mTOR signallingtargeted therapies.

Conclusion: This study suggests S6K2 and 4EBP1 as important factors for breast tumourigenesis, interplaying withhormone receptor signalling. We propose S6K2 and 4EBP1 as new potential clinical markers for prognosis andendocrine therapy response in breast cancer.

* Correspondence: [email protected] of Clinical and Experimental Medicine, Division of Oncology,Linköping University, County Council of Östergötland, Linköping SE-58185,SwedenFull list of author information is available at the end of the article

© 2013 Karlsson et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.

mailto:[email protected]

http://creativecommons.org/licenses/by/2.0

Karlsson et al. Breast Cancer Research 2013, 15:R96 Page 2 of 14http://breast-cancer-research.com/content/15/5/R96

IntroductionThe outcome of breast cancer patients has been consid-erably improved in recent years, as a result of early diag-nosis and improved treatment regimens; however, breastcancer remains a leading cause of malignancy-associateddeath among women worldwide. Traditionally, breast can-cers have been classified into prognostically meaningfulgroups based on clinical features and histopathologicalfindings, but it is increasingly evident that cellular andmolecular characteristics are of significant importance.Oestrogen receptor alpha (ER), expressed in 70 to 80%

of breast cancers, is a standard biomarker for predictionof response to endocrine treatment. However, significantproportions of ER-positive tumours are resistant to en-docrine therapy, either de novo or acquired, and morespecific biomarkers as well as new therapeutic targetsfor endocrine-resistant tumours are needed. Suggestedmechanisms of endocrine resistance include loss of ERexpression or expression of truncated ER isoforms, post-translational modification of the ER, deregulation ofcofactors, or overstimulation of tyrosine kinase receptorgrowth signalling pathways [1].The serine/threonine kinase mammalian/mechanistic

target of rapamycin (mTOR) is assumed to be a criticaleffector for several cellular functions deregulated in cancer[2]. mTOR exists in two cellular complexes, referred to asmTORC1 and mTORC2. In response to growth factors,hormones, nutrients, hypoxia and energy/ATP, mTORC1regulates cell growth, proliferation and metabolism throughtranslational control of essential proteins. The most well-known substrates of mTORC1 are the 4E-binding protein1 (4EBP1) and the p70 ribosomal S6 kinases 1 and 2 (S6K1and S6K2), which are involved in regulation of the transla-tional machinery [2]. Two major regulators of mTORC1function, the rat sarcoma oncogene/mitogen-activated pro-tein kinase and phosphatidylinositol-3-kinase (PI3K)/AKTsignalling pathways are constitutively activated in manycancers; however, the mechanisms behind mTORC2 acti-vation are less known. mTORC2 has been shown to bephosphorylated and activated in response to growth fac-tors, but the intracellular pathways remain to be unrav-elled. The complex has been implicated in cytoskeletaldynamics, through activation of Rho GTPases and PKCα,but also in regulation of AKT through direct phoshoryla-tion of Ser473, thereby promoting its activation [2].The most frequently altered intracellular growth sig-

nalling pathway in breast cancer is PI3K/AKT/mTOR,which is suggested as a key driver of proliferation andsurvival, particularly in ER-positive tumours. PI3K/AKT/mTOR and ER are implicated in a bidirectional cross-talk, in which intracellular signalling pathways stimulategenomic ER signalling through phosphorylation and ac-tivation of the receptor and its cofactors. In addition,oestrogen stimulation of breast cancer cells immediately

upregulates intracellular kinase signalling, suggesting non-genomic signalling through cytoplasmic or membranebound ER to be involved in activation of PI3K/AKT/mTOR signalling [3]. Targeting mTOR has emerged asa new promising treatment strategy for several malig-nancies and recent data indicate that combining endo-crine therapy in breast cancer with mTOR inhibitors iseffective [4,5].Studies have indicated the importance of alterations in

factors downstream of mTOR for the development ofmalignancy. S6K1 as well as S6K2 have been shown tobe upregulated in breast cancer [6]. The genes RPS6KB1(S6K1) and RPS6KB2 (S6K2) are situated in the chromo-somal regions 17q21-23 and 11q13, which are commonlyamplified in several malignancies. S6K1 amplification andS6K1 protein overexpression have previously been associ-ated with a worse outcome in breast cancer [7-9]. Wehave also recently shown that S6K2 is amplified and over-expressed in breast tumours, and the results indicated thatS6K1 and S6K2 amplification may have prognostic signifi-cance independent of the neighbouring oncogenes ERBB2and CCND1 [8].Phosphorylation of 4EBP1 by mTORC1 promotes dis-

sociation of 4EBP1 from EIF4E, enabling EIF4E to induceprotein translation. Consequently, phosphorylated 4EBP1(p4EBP1) has been generally accepted as a marker of acti-vated mTOR signalling and high levels in tumours havebeen associated with a worse outcome in several ma-lignancies, whereas nonphosphorylated 4EBP1 has beenconsidered a tumour suppressor [10]. However, the geneencoding 4EBP1 is located at the chromosomal region8p12, which is commonly amplified in breast cancer, andin a recent study gene amplification and high mRNAlevels of 4EBP1 were shown to indicate a poor prognosis[11]. This suggests that 4EBP1 may have an active role incarcinogenesis. Accordingly, 4EBP1 has also been shownto bind and stabilise mTORC1, promoting activation ofthe signalling pathway [12].The mTORC1/S6K/4EBP1 pathway is a major regulator

of protein synthesis by phosphorylating several factors inthe translational initiation complex, and is thus consideredas mainly acting in the cytoplasm [13]. However, recentstudies have shown that mTOR as well as the S6 kinasesand 4EBP1 can shuttle between the cytoplasm and the nu-cleus, and are indicated to be involved in regulation oftranscription [14-17].The present aim was to further investigate the signifi-

cance of 4EBP1 together with S6K1 and S6K2 in breastcancer, in a study encompassing five different cohorts ofpatients. We showed that S6K2 and 4EBP1 have a corre-lated mRNA expression, and that high levels of S6K2and/or 4EBP1 were associated with a poor prognosis, inde-pendently of other classical prognostic markers. Further-more, high cytoplasmic levels of 4EBP1 protein predicted a


poor prognosis, whereas 4EBP1 expression, regardlessof cellular location, was associated with a decreasedbenefit from endocrine treatment, suggesting a new rolefor 4EBP1 in hormone receptor signalling. This studyestablishes the mTOR effectors 4EBP1 and S6K2, asnew potential clinical markers in breast cancer diagnos-tics and treatment prediction.

MethodsThe study encompasses two cohorts from the rando-mised adjuvant Stockholm tamoxifen trials, referred toas Stockholm 2 and Stockholm 3. In addition, three pub-lically available datasets were used to confirm the results.The design of the present study and the results presenta-tion are in line with the Reporting Recommendations forTumour Marker Prognostic Studies guidelines [18].

Patients in the randomised Stockholm tamoxifen trialsThe Stockholm 2 and Stockholm 3 cohorts consist ofpostmenopausal breast cancer patients enrolled in ran-domised adjuvant studies between November 1976 andApril 1990. Study designs and long-term follow-up datawere previously reported in detail [19,20]. Briefly, pa-tients in the Stockholm 2 cohort had positive lymphnodes and/or a tumour diameter exceeding 30 mm,whereas the Stockholm 3 cohort consisted of breast can-cer patients with a tumour diameter ≤30 mm and nolymph node involvement. All patients were randomisedto receive tamoxifen for 2 years or no endocrine treat-ment. Patients in the Stockholm 2 cohort were furtherrandomised to postoperative radiotherapy or cyclophos-phamide–methotrexate–5-fluorouracil-based chemother-apy. Most of the patients in the tamoxifen arm, if diseasefree after 2 years, were then randomised to receive tam-oxifen for 3 years more or no further adjuvant treatment.Patient flow through the study is presented in Additionalfile 1: Figure S1 and in Additional file 2. Clinicopatho-logical data can be found in Additional file 3. For thepresent study, 93 and 912 tumour samples were avail-able from the Stockholm 2 and Stockholm 3 cohorts, re-spectively. Tumour characteristics and treatments werecomparable with the original cohort.Ethical approval for the Stockholm 2 and Stockholm 3

cohorts was from Karolinska Institute Ethics Council (DnrKI 97–451 with amendment 030201). Retrospective studiesof biomarkers were approved by the local ethics board atthe Karolinska Institute, Stockholm, Sweden. Further needfor patient consent was waived by the ethical review board.

RNA extraction and real-time polymerase chain reaction(Stockholm 2)Fresh-frozen tumour tissue, estimated to contain >50%cancer cells, was homogenised with a microdismembrator(B. Braun Melsungen AG, Germany) or a tissue lyser

(Qiagen Hilden, Germany) and total RNA was isolated withthe mirVana™ miRNA isolation kit (Ambion Life Technol-ogy, Grand Island, NY, USA), according to instructions pro-vided by the manufacturers. Purified RNA was dissolved innuclease-free water with addition of RNAsin Ribonucleaseinhibitor (Promega Madison, WI, USA) and was storedat −70°C. RNA integrity numbers and concentrations wereassessed with an Agilent 2100 Bioanalyser (Agilent Biosys-tems Santa Clara, CA, USA). Only samples with RNA in-tegrity numbers ≥5 were included in the analysis.Reverse transcription was performed using the high-

capacity cDNA reverse transcription kit (Applied BiosystemsFoster City, CA, USA) with 200 ng total RNA in reactions of20 μl according to the manufacturer’s instructions. mRNAexpression of S6K1, S6K2 and 4EBP1 was quantified withfast real-time polymerase chain reaction (PCR) using anABI Prism 7900ht (Applied Biosystems). TaqMan assays(Applied Biosystems) for S6K1 (Hs00177357_m1), S6K2(Hs00177689_m1), 4EBP1 (Hs0060705_m1) and the en-dogenous controls β-actin (ACTB; part number 4310881E)and peptidylprolyl isomerase A (cyclophilin A) (PPIA; partnumber 4333763 F) were handled according to the manu-facturer’s instructions. Quantitative PCR was performed induplicate with 10 μl reaction volume in 1× TaqMan fast uni-versal master mix (Applied Biosystems) using the followingthermal conditions: 95°C for 20 seconds; 40 cycles of 95°Cfor 1 second, and 60°C for 20 seconds. To confirm specifi-city, reactions without reverse transcriptase as well as notemplate controls were included on each plate. The meanvalue was taken from the duplicates and relative expressionwas calculated with the ΔΔCt method, using SKBR3 cDNAas the calibrator. For the two endogenous controls, an aver-age value for each sample was used. For correlation analyses,expression levels of the genes were divided into four groupsbased on the quartiles. In the survival analyses, the upperquartile was considered as high expression and the remaininglevels as low expression, if nothing else is specified.

Tissue microarray preparation and immunohistochemicalanalysis (Stockholm 3)The protein expressions of total 4EBP1 and 4EBP1 phos-phorylated at Serine 65 (hereafter referred to as 4EBP1and p4EBP1, respectively) were evaluated in the Stockholm3 cohort by immunohistochemical (IHC) staining of tissuemicroarrays. Core needle biopsies from paraffin-embeddedtissues were reembedded in new paraffin blocks and theblocks were cut into 4 μm sections and mounted on frost-coated slides. The slides were deparaffinised in xylene andrehydrated in decreasing concentrations of ethanol, andantigen retrieval was performed in citrate buffer (pH 6.0)in a pressure cooker with the default program 125°C for30 seconds followed by 90°C for 10 seconds at a pressureof 23 to 25 psi. Endogenous peroxidases were blockedwith 3% H2O2 in MeOH for 5 minutes, and protein block


X0909 (Dako Glostrup, Denmark) was applied for 10 mi-nutes to reduce unspecific binding. The slides were incu-bated with primary antibodies for 4EBP1 (53H11/#9644,diluted 1:100; Cell Signaling Danvers, MA, USA) orp4EBP1_S65 (174A9/#9456, diluted 1:100; Cell Signaling)overnight at 4°C. Secondary antibody (EnVision™;Dako) was applied for 30 minutes at room temperature.For visualisation, the slides were incubated in 3,3′-diami-nobenzidine hydrochloride/H2O2 for 8 minutes at roomtemperature and in darkness, and counterstained withhaematoxylin for 1 minute at room temperature and indarkness. Representative images of the stainings werephotographed at 40× magnification using an Olympus SC20digital camera (Olympus Järfälla, Stockholm, Sweden) con-nected to a Leica LB30T microscope (Leica Microsys-tems Kista, Stockholm, Sweden) (Additional file 1:Figures S2 and S3). Phospho-specificity for p4EBP1_S65was evaluated with lambda phosphatase (New EnglandBiolabs Ipswitch, UK) according to manufacturer’s in-structions (Additional file 1: Figure S3). Protein specificityof the 4EBP1 antibodies was validated with western blot,by us and others [16,21] (Additional file 1: Figures S2 and S3).Cytoplasmic and nuclear intensity of the stainings was eval-uated by two independent observers, according to the levelsdepicted in Additional file 4. In the survival analyses, a high4EBP1 expression was defined as strong cytoplasmic or nu-clear staining, whichever indicated. The variable 4EBP1cy-toplasm ≥ nucleus was defined as a cytoplasmic stainingstronger than or equal to the nuclear staining detected.

Evaluation of other clinicopathological variablesER expression was determined at the time of diagnosis,before 1988 using isoelectric focusing and after that withquantitative enzyme immunoassay [19,20]. In the Stockholm3 cohort, where tissue microarrays were available, the ERand progesterone receptor (PgR) status was further de-termined retrospectively by IHC using the Ventanaautomated slide stainer (Ventana Medical SystemsTucson, Arizona) with monoclonal Ventana CONFIRMmouse primary ER and PgR antibodies [22]. The cutofflevel for ER and PgR positivity was >10% stained nuclei or,when IHC data were not available, 0.05 fmol/μg DNA.Isoelectric focusing/enzyme immunoassay and IHC datahave been shown to be comparable [23]. In the Stockholm2 cohort, human epidermal growth factor receptor 2(HER2) protein was quantified retrospectively by flow cy-tometry [24] and HER2 amplification was determinedwith quantitative real-time PCR [25]. HER2 protein ex-pression in the Stockholm 3 cohort was evaluated withIHC as described elsewhere [26], whereas tumour gradewas evaluated retrospectively according to the Notting-ham system [22]. In the Stockholm 2 cohort, S-phase frac-tion was previously determined by flow cytometry [27].

Extraction of DNA from fresh-frozen tissue and analysis ofthe S6K1 and S6K2 gene copy number were described else-where [8,11]. Analyses of mutations in PIK3CA as well asprotein expression of pAKT_S473 in the Stockholm 2 co-hort were reported earlier [28,29]. In the Stockholm3 cohort, the S6K2, pAKT_S473 and pmTOR_S2448IHC stainings have also been described previously [8,30].

Public datasets (van de Vijver, Uppsala, Karolinska)Public available datasets encompassing preprocessed mRNAexpression data were downloaded for three cohorts, furtherreferred to as the van de Vijver cohort (n = 295) [31], theUppsala cohort (n = 236) [NCBI/GEO:GSE3494] and theKarolinska Institute cohort (n= 159) [NCBI/GEO:GSE1456].Patient flow is overviewed in Additional file 2. The patientcharacteristics are briefly described in Additional file 3and were previously presented in detail, as was the dataprocessing method [9,32,33].

Statistical analysisAssociations between different variables were assessed bySpearman’s rank-order correlation. The Kaplan–Meierproduct limit method was used to estimate the cumulativeprobabilities of distant recurrence-free survival or breastcancer-specific survival, and differences between the curveswere evaluated with the log-rank test or Gehan’s test formultiple groups. For univariate and multivariate analysisof event rates, as well as interaction analysis, Cox propor-tional hazard regression was used.In the interaction test, the Cox model included the

variables ‘tamoxifen treatment’ and ‘4EBP1 expression’and the interaction variable ’tamoxifen treatment × 4EBP1expression’. All statistical analyses were performed withStatistica 9.0 (Statsoft, Inc. Uppsala, Sweden) and P <0.05was considered statistically significant, with exception ofthe correlation analyses where P <0.01 was applied tocompensate for multiple testing.

ResultsGene amplifications of S6K1 and S6K2 are associated withhigh levels of corresponding mRNA4EBP1, S6K1 and S6K2 mRNA levels were quantified in93 tumours from the Stockholm 2 cohort. S6K1 and S6K2gene amplification was previously determined with real-time PCR in 206 and 207 breast tumour samples, respect-ively [8]. There was a significant correlation between genecopy number and mRNA levels for both genes (S6K1:Spearman R = 0.30, P = 0.007; S6K2: Spearman R = 0.43,P = 0.0001). An increased gene copy number was almostalways accompanied by high mRNA levels, but highmRNA levels could be detected in additional samples,independent of gene copy status (Additional file 5).


4EBP1 mRNA is frequently coexpressed with S6K2, butnot with S6K1In a previous study encompassing 29 of the Stockholm 2patients, S6K2 and 4EBP1 were found to be coamplifiedand expression levels for the corresponding mRNAs werecorrelated [11]. In line with this finding, when consideringall 93 patients in the present study, S6K2 and 4EBP1mRNA levels were significantly correlated (Spearman R =0.32, P = 0.0018). There was no correlation between S6K1and 4EBP1 mRNA levels (Spearman R = −0.0017, P = 0.99;Figure 1). S6K1 mRNA was positively correlated with ERstatus (Spearman R = 0.39, P = 0.00009; Additional file 5).There was also an inverse association between high S6K1mRNA levels and HER2 amplification/protein levels aswell as high S-phase fraction (Additional file 5). A correl-ation between S6K2 and 4EBP1 mRNA expression couldbe confirmed in the three public cohorts, whereas S6K1and 4EBP1 mRNA levels were associated with high sig-nificance in the Karolinska cohort only (Figure 1). Theassociation between S6K1 and ER status in Stockholm 2could not be detected in the other cohorts (data notshown).

High mRNA levels of S6K2 and 4EBP1 are associated withan adverse outcome in four breast cancer cohortsS6K1, S6K2 and 4EBP1 gene amplification have earlierbeen connected to a worse prognosis in breast cancer.At the mRNA level, S6K2 and 4EBP1 remained inde-pendent prognostic factors in the Stockholm 2 cohort,whereas this could not be seen for S6K1 (Figure 2a,b,c).

S6K1mRNA

4EBP1mRNA

S6K2mRNA

Stockholm 2van de Vijver

KarolinskaUppsala

*

***

***

***

*******

*p<0.05, **p<0.01, ***p<0.001

Figure 1 Associations between p70 ribosomal S6 kinase 1 and2 and 4E-binding protein 1 mRNA expression. Spearman’s rank-order correlation evaluating associations between p70 ribosomal S6kinase 1 (S6K1), p70 ribosomal S6 kinase 2 (S6K2) and 4E-bindingprotein 1 (4EBP1) mRNA expression (continuous values) in fourbreast cancer cohorts.

For 4EBP1, the prognostic value was especially pronouncedin the ER-positive subgroup (Figure 2d). A combinationvariable of high S6K2 and/or 4EBP1 mRNA was a signifi-cant independent prognostic factor, and the worst outcomecould be seen in the group with the highest levels of bothS6K2 and 4EBP1 (Figure 3a). The prognostic value of S6K1,S6K2 and 4EBP1 mRNA was further analysed in the threepublic cohorts (Additional file 1: Figures S4, S5 and S6).4EBP1 remained an independent prognostic factor in thevan de Vijver and Karolinska cohorts. S6K2 was also signifi-cantly associated with clinical outcome in the Karolinskacohort and, when divided into two groups based on themedian, this was also true in the van de Vijver cohort. Inthe Uppsala cohort, S6K2 and 4EBP1 remained prognosticfactors in the univariate analysis, whereas the multivariateanalyses did not reach significance. S6K1 was significantlyassociated with a worse outcome in the van de Vijver co-hort only. The combined variable S6K2 and/or 4EBP1mRNA was confirmed as a significant prognostic factor,related to poor outcome, in the van de Vijver and Karo-linska cohorts, and a borderline significance was seen inthe Uppsala cohort (Figure 3b,c,d).There was a significant correlation between high S6K2

and/or 4EBP1 to grade in the Uppsala and Karolinskacohorts as well as to the proliferation marker cyclin A2

in the van de Vijver cohort. In the Stockholm 2 cohort,the correlation between S6K2 and/or 4EBP1 and high S-phase fraction reached borderline significance. High S6K2and/or 4EBP1 was primarily seen in ER/PgR-negative tu-mours in the van de Vijver and Uppsala cohorts and thesame tendency could be seen in the Karolinska cohort.High S6K2 and/or 4EBP1 was also significantly associatedwith large tumour size in the Uppsala material (Table 1).

Clinicopathological characteristics of 4EBP1-expressingtumours are dependent on the cellular localisation of theproteinProtein expression of 4EBP1 and p4EBP1 could be analysedin 739 and 768 tumours, respectively, in the Stockholm 3cohort. 4EBP1 and p4EBP1 were detected in both the nu-cleus and the cytoplasm of the tumour cells (Additionalfile 4). Correlations between 4EBP1 and p4EBP1 proteinexpression and clinicopathological factors are described inAdditional file 6. Strong cytoplasmic 4EBP1 and p4EBP1expression was associated with high-grade and HER2-positive tumours and also with large tumour size. Nuclearp4EBP1 was associated with small, low-grade tumours.Nuclear and cytoplasmic p4EBP1 were significantly cor-related with pAKT expression in the respective com-partments. There was no significant correlation betweenpmTOR and p4EBP1 or 4EBP1. Both p4EBP1 and cyto-plasmic 4EBP1 were significantly associated with S6K2protein expression (Additional file 6).

S6K1 mRNAStockholm 2, all patients

0 2 4 6 8 10 12 14

Years

0,0

0,2

0,4

0,6

0,8

1,0

Bre

ast c

ance

r su

rviv

al

S6K1 mRNA high (n=23)

S6K1 mRNA low (n=70)

p=0.79

S6K2 mRNAStockholm 2, all patients

0 2 4 6 8 10 12 14

Years

0,0

0,2

0,4

0,6

0,8

1,0

Bre

ast c

ance

r su

rviv

al

S6K2 mRNA low (n=69)

S6K2 mRNA high (n=24)

p=0.00090

4EBP1 mRNAStockholm 2, all patients

0 2 4 6 8 10 12 14

Years

0,0

0,2

0,4

0,6

0,8

1,0

Bre

ast c

ance

r su

rviv

al 4EBP1 mRNA low (n=69)

4EBP1 mRNA high (n=24)

p=0.12

4EBP1 mRNAStockholm 2, ER+

0 2 4 6 8 10 12 14

Years

0,0

0,2

0,4

0,6

0,8

1,0

Bre

ast c

ance

r su

rviv

al

4EBP1 mRNA high (n=17)

4EBP1 mRNA low (n=55)

p=0.024

a) b)

c) d)

Figure 2 Breast cancer survival and distant recurrence-free survival in the Stockholm 2 patient cohort in relation to S6K1, S6K2and 4EBP1 mRNA levels. Kaplan–Meier curves and multivariate Cox regression of breast cancer survival (BCS) and distant recurrence-freesurvival (DRFS) in the Stockholm 2 patient cohort, in relation to (a) p70 ribosomal S6 kinase 1 (S6K1) mRNA, (b) p70 ribosomal S6 kinase 2(S6K2) mRNA, (c) 4E-binding protein 1 (4EBP1) mRNA, and (d) 4EBP1 mRNA in the oestrogen receptor alpha (ER)-positive subgroup. TheCox analysis included the following variables: adjuvant chemotherapy treatment, endocrine treatment, lymph node status, tumour size, andER status. CI, confidence interval; HR, hazard ratio.


p4EBP1 and 4EBP1 protein expression are independentprognostic factors in breast cancerHigh tumour levels of p4EBP1 have earlier been associ-ated with poor outcome in breast cancer and other malig-nancies. For systemically untreated patients, in the presentstudy, strong cytoplasmic p4EBP1 staining remained anindependent prognostic factor, predicting decreased dis-tant recurrence-free survival and poor breast cancer sur-vival (Figure 4a). In contrast, nuclear p4EBP1 did not

correlate with prognosis (Figure 4b), while strong nuclear4EBP1 staining indicated good prognosis (Figure 4c), andthis was especially evident in the PgR-positive subgroup(Figure 4d). No prognostic significance could be seen forcytoplasmic 4EBP1 (data not shown), but the variable4EBP1cytoplasm ≥ nucleus was an independent prognosticfactor, predicting increased risk of distant recurrence andbreast cancer death (Figure 4e), especially among patientswith PgR-expressing tumours (Figure 4f).

S6K2 and/or 4EBP1 mRNAStockholm 2, all patients

0 2 4 6 8 10 12 14

Years

0,0

0,2

0,4

0,6

0,8

1,0

Bre

ast c

ance

r su

rviv

al

p=0.0021

S6K2 and 4EBP1 high (n=9)

S6K2 or 4EBP1high (n=30)

S6K2 and 4EBP1 low (n=54)

S6K2 and/or 4EBP1 mRNAvan de Vijver, all patients

0 2 4 6 8 10 12 14 16 18 20

Years

0,0

0,2

0,4

0,6

0,8

1,0

Bre

ast c

ance

r su

rviv

al

p=0.000010


S6K2 or 4EBP1 high (n=88)


S6K2 and/or 4EBP1 mRNAUppsala, all patients

0 2 4 6 8 10 12 14 16

Years

0,0

0,2

0,4

0,6

0,8

1,0

Bre

ast c

ance

r su

rviv

al

p=0.000050

S6K2 and 4EBP low (n=158)

S6K2 or 4EBP1 high (n= 62)


S6K2 and/or 4EBP1 mRNAKarolinska, all patients

0 1 2 3 4 5 6 7 8 9 10

Years

0,0

0,2

0,4

0,6

0,8

1,0

Bre

ast c

ance

r su

rviv

al

p=0.00028


S6K2 or 4EBP1 high (n=46)


a) b)

c) d)

Figure 3 Breast cancer survival and distant recurrence-free survival in relation to high S6K2 and/or 4EBP1 mRNA in four cohorts.Kaplan–Meier curves and multivariate Cox regression of breast cancer survival (BCS) and, when available, of distant recurrence-free survival (DRFS)in relation to high p70 ribosomal S6 kinase 2 (S6K2) and/or 4E-binding protein 1 (4EBP1) mRNA in four cohorts: (a) Stockholm 2, (b) van de Vijver,(c) Uppsala, and (d) Karolinska. The variable S6K2 and/or 4EBP1 was defined in two groups as highest quartile mRNA expression of: neither S6K2nor 4EBP1 (=0), or of S6K2 and/or 4EBP1 (=1). The Cox analysis also included the following variables: adjuvant chemotherapy treatment,endocrine treatment, lymph node status, tumour size (with exception of van de Vijver), and oestrogen receptor alpha status. CI, confidenceinterval; HR, hazard ratio.


High cytoplasmic protein levels of 4EBP1 predict adecreased benefit from endocrine treatmentUpregulation of the AKT/mTOR pathway has been im-plicated as one mechanism behind endocrine resistance.In the Stockholm 3 cohort, the outcome among patientswith ER-positive/PgR-positive tumours treated with tam-oxifen was evaluated in relation to 4EBP1 protein expres-sion in different compartments (Figure 5). This analysisconfirmed cytoplasmic 4EBP1 to be predictive of poor clin-ical outcome in the tamoxifen-treated ER-positive /PgR-

positive group (Figure 5c), as well as the variable 4EBP1cytoplasm ≥ nucleus (Figure 5e). In addition, cytoplasmicp4EBP1 (Figure 5a) was shown borderline significant in re-lation to a poor prognosis in this patient group . Nuclearp4EBP1 or nuclear 4EBP1 was not related to outcome aftertamoxifen treatment (Figure 5b,d). In a subsequent ana-lysis, the benefit from tamoxifen was compared betweenpatients with ER-positive/PgR-positive tumours expressinglow or high cytoplasmic levels of p4EBP1 or 4EBP1. Tam-oxifen treatment was associated with a strongly reduced

Table 1 Association between high S6K2 and/or 4EBP1 mRNA and clinicopathological parameters in the four cohorts

S6K2 and/or 4EBP1 mRNA (Spearman R, P value)

Stockholm 2 Van de Vijver Uppsala Karolinska

(n = 93) (n = 295) (n = 251) (n = 159)

Lymph node status R = −0.11 R = 0.07 R = 0.31 R = −0.07

P = 0.25 P = 0.21 P = 0.000001 P = 0.36

Tumour size R = −0.12 N/A R = 0.32 R = 0.16

P = 0.23 P < 0.0000001 P = 0.040

Grade/proliferationa R = 0.23 R = 0.42 R = 0.45 R = 0.41

P = 0.034 P < 0.0000001 P <0.0000001 P = <0.0000001

ER status R = −0.01 R = −0.31 R = −0.26 R = −0.11

P = 0.91 P <0.0000001 P = 0.000046 P = 0.15

PgR status (protein) N/A N/A R = −0.27 N/A

P= 0.000010

PgR mRNA N/A R = −0.29, R = −0.30 R = −0.16

P <0.0000001 P= 0.000002 P = 0.048

Spearman’s rank-order correlation evaluating the association between high S6K2 and/or 4EBP1 mRNA and clinicopathological parameters in the four cohorts. Thevariable S6K2 and/or 4EBP1 was defined in three groups as highest quartile mRNA expression of: neither S6K2 nor 4EBP1 (=0), either S6K2 or 4EBP1 (=1), or S6K2 and4EBP1 (=2). 4EBP1, 4E-binding protein 1; ER, oestrogen receptor alpha; PgR, progesterone receptor; S6K2, p70 ribosomal S6 kinase 2.aStockholm 2, S-phase fraction; van de Vijver, cyclin A2 mRNA expression; Uppsala and Karolinska, Elston grade. N/A, not available.


risk of distant recurrence in the group of patients withER-positive/PgR-positive tumour and low cytoplasmic 4EBP1(distant recurrence-free survival: hazard ratio (95% confi-dence interval) = 0.19 (0.09 to 0.42), P= 0.00003; Figure 6a),whereas no significant benefit from tamoxifen could be seenin the 4EBP1 high cytoplasmic group (distant recurrence-free survival: hazard ratio (95% confidence interval) = 0.60(0.30 to 1.23), P = 0.17; Figure 6b). The difference intreatment benefit between the groups with low and highcytoplasmic 4EBP1 was significant (test for interaction,P = 0.034). The interaction test concerning cytoplasmicp4EBP1 did not reach significance (data not shown).

DiscussionThe role of mTOR signalling in cancer development, pro-gression and as a potential treatment target is increasinglyevident. In this study, we highlight the clinical importanceof factors downstream of mTOR, and show that mRNAexpression of S6K2 and 4EBP1 are correlated and signifi-cantly related to poor outcome in four independent breastcancer cohorts. This is the first study showing high 4EBP1mRNA, independent of phosphorylation status, and cyto-plasmic protein levels to be associated with poor progno-sis in breast cancer. Furthermore, high 4EBP1 proteinlevels predicted less benefit from the endocrine treatmenttamoxifen, indicating interactions with hormone receptorsignalling. This suggests that the mTOR effectors S6K2and 4EBP1 may be used as prognostic indicators and fortreatment prediction.The S6 kinases are frequently upregulated in breast

cancer, and associated with a poor outcome [6,34]. In

the present study, we could show a correlation betweengene amplification and increased mRNA levels for S6K1,S6K2 as well as seen previously for 4EBP1 [11]. Tumourswith amplification of these genes had high levels of thecorresponding mRNA; however, high mRNA expressionwas also in some cases seen in tumours with normal genecopy numbers. Recently, S6K1 was described as a transcrip-tional target of the ER [35]. Here, there is a correlation be-tween ER and S6K1 mRNA levels in the Stockholm 2cohort, suggesting that ER expression could be one mech-anism behind S6K1 upregulation in breast tumours. How-ever, S6K1 gene amplification in Stockholm 2 was in aprevious study correlated with HER2 positivity rather thanER expression [8], probably as a consequence of the local-isation of the S6K1 gene in proximity of the ERBB2 geneat 17q. It is evident that, although amplification and ex-pression of these genes are tightly accompanied, theseevents are not identical. Gene amplification probably re-flects the contribution of several genes in the amplicons,and the feature of expression is highly dependent on thecellular localisation of the proteins.The previously implicated associations between S6K2

and 4EBP1 [11] were further confirmed in this study,and could be seen in several independent and clinicallydifferent patient materials. High S6K2 and/or 4EBP1 mRNAwas associated with poor outcome in all investigated co-horts, which may reflect a possible synergy between S6K2and 4EBP1 in promoting tumourigenesis. p4EBP1 hasbeen shown to predict a poor prognosis in several cancertypes [36-39] and the protein was recently described as akey funnel factor in carcinogenesis [10]. In general, p4EBP1

4EBP1 protein cytoplasm > nucleusStockholm 3, systemically untreated patients

PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l

4EBP1cyt > nucleus (n=22)

4EBP1cyt < nucleus (n=142)

p=0.0053

p4EBP1 protein nuclearStockholm 3, systemically untreated patients

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l

p4EBP1n low (n=142)

p4EBP1n high (n=236)

p=0.82

p4EBP1 protein cytoplasmStockholm 3, systemically untreated patients

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l

p4EBP1cyt high (n=118)

p=0.040

p4EBP1cyt low (n=257)

4EBP1 protein nuclearStockholm 3, systemically untreated patients

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l

4EBP1n high (n=125)

4EBP1n low (n=233)

p=0.086

4EBP1 protein cytoplasm > nucleusStockholm 3, systemically untreated patients

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l



p=0.018

c)

b)a)

d) 4EBP1 protein nuclearStockholm 3, systemically untreated patients

PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0D

ista

nt r

ecur

renc

e-fr

ee s

urvi

val

p=0.017

4EBP1n high (n=54)

4EBP1n low (n=110)

e) f)

Figure 4 Survival among systemically untreated patients in Stockholm 3 in relation to p4EBP1_S65 and 4EBP1 protein. Kaplan–Meiercurves and multivariate Cox regression of distant recurrence-free survival (DRFS) and breast cancer survival (BCS) among systemically untreatedpatients in the Stockholm 3 cohort in relation to (a) p4EBP1_S65 cytoplasmic protein, (b) p4EBP1_S65 nuclear protein, (c) 4E-binding protein 1(4EBP1) nuclear protein, (d) 4EBP1 nuclear protein in the progesterone receptor (PgR)-positive subgroup, (e) 4EBP1 protein cytoplasm≥ nucleus,and (f) 4EBP1 protein cytoplasm≥ nucleus in the PgR-positive subgroup. The Cox analysis included the following variables: adjuvant tamoxifentreatment, tumour size, Nottingham histological grade, oestrogen receptor alpha, and human epidermal growth factor receptor 2 (HER2) status.*For the PgR-positive/BCS multivariate analyses, HER2 was divided into four groups due to few events. CI, confidence interval; HR, hazard ratio.


4EBP1 protein cytoplasm > nucleusStockholm 3, tamoxifen treated

ER+, PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l



p=0.011

4EBP1 protein cytoplasmStockholm 3, tamoxifen treated

ER+, PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l

4EBP1cyt high (n=60)

p=0.0010

4EBP1cyt low (n=136)

4EBP1 protein nucleusStockholm 3, tamoxifen treated

ER+, PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0D

ista

nt r

ecur

renc

e-fr

ee s

urvi

val 4EBP1nucleus low (n=123)

4EBP1 nucleus high (n=73)

p=0.63

p4EBP1 protein cytoplasmStockholm 3, tamoxifen treated

ER+, PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l

p4EBP1cyt low (n=158)

p4EBP1cyt high (n=37)

p=0.11

p4EBP1 protein nucleusStockholm 3, tamoxifen treated

ER+, PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l p4EBP1 nucleus high (n=112)

p4EBP1 nucleus low (n=83)

p=0.44

e)

a) b)

c) d)

Figure 5 Distant recurrence-free survival among patients in Stockholm 3 with ER-positive/PgR-positive tumours treated with tamoxifenin relation to p4EBP1_S65 and 4EBP1 protein expression. Kaplan–Meier curves and multivariate Cox regression of distant recurrence-free survival(DRFS) among patients in the Stockholm 3 cohort with ER-positive/PgR-positive tumours, treated with tamoxifen, in relation to (a) p4EBP1_S65cytoplasmic protein, (b) p4EBP1_S65 nuclear protein, (c) 4E-binding protein 1 (4EBP1) cytoplasmic protein, (d) 4EBP1 nuclear protein, and(e) 4EBP1protein cytoplasm≥ nucleus. The Cox analysis included the following variables: tumour size, Nottingham histological grade, and humanepidermal growth factor receptor 2 (HER2) status. *For the 4EBP1 analysis, HER2 was divided into four groups due to few events. BCS, breastcancer-specific survival; CI, confidence interval; ER, oestrogen receptor alpha; HR, hazard ratio; PgR, progesterone receptor.


4EBP1 protein cytoplasm lowER+, PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l

Tam (n=136)

no Tam (n=105)

p<0.000001

4EBP1 protein cytoplasm highER+, PgR+

0 5 10 15 20 25 30

Years

0,0

0,2

0,4

0,6

0,8

1,0

Dis

tant

rec

urre

nce-

free

sur

viva

l

Tam (n=60)

no Tam (n=53)

p=0.16

a) b)

Figure 6 Distant recurrence-free survival for ER-positive/PgR-positive breast cancer patients in Stockholm 3 treated or not withtamoxifen in relation to 4EBP1 cytoplasmic protein. Distant recurrence-free survival (DRFS) for breast cancer patients in the Stockholm 3cohort treated with tamoxifen (Tam) versus no tamoxifen (no Tam) in the ER-positive/PgR-positive subgroup in relation to (a) low and (b) highcytoplasmic 4E-binding protein 1 (4EBP1) protein levels. ER, oestrogen receptor alpha; PgR, progesterone receptor.


has been considered a marker of mTORC1 signalling andactivation of the translational machinery. However, thereare indications that 4EBP1 could also play a more activerole in tumour progression. In this study, cytoplasmicp4EBP1 was confirmed as a prognostic factor; however,4EBP1 and 4EBP1 mRNA were also associated with highgrade and a poor outcome. The gene encoding 4EBP1 islocated in the chromosomal region 8p12, which is fre-quently amplified in breast cancer. Amplification of 8p12was associated with high 4EBP1 mRNA levels, suggesting4EBP1 as a potential oncogene, and amplification of the4EBP1 gene is one possible mechanism behind its overex-pression in tumours. Another suggested pathway is throughcMyc-dependent transcription, and amplification or inducedexpression of cMyc has been shown to promote cMycbinding to the 4EBP1 gene and increase its expression.This in turn leads to inhibition of autophagy and rapamy-cin resistance [40]. Unfortunately, we have not been ableto study the possible relation between 4EBP1 mRNAlevels and its corresponding protein expression. A recentreview on the issue of regulation of protein expressionand its relation to mRNA levels conclude that the abun-dance of mRNA in general highly reflects the ability to de-tect protein expression in cells [41]. High mRNA levels of4EBP1 as well as high cytoplasmic protein levels are bothrelated to a high proliferation and a poor prognosis in thedifferent materials investigated. One could therefore spe-culate that high mRNA levels may reflect increased cy-toplasmic protein levels rather than nuclear, perhaps as aresult of increased nuclear-cytoplasmic shuttling in prolifer-ating cells, although the mechanisms behind this are unclear.Interestingly, the prognostic value of 4EBP1 seems to

be dependent on the cellular location of the protein.

Nuclear expression was related to a better outcome, in-dicating that 4EBP1 plays divergent roles in different cel-lular compartments. A previous study estimated thatapproximately 30% of the 4EBP1 expressed in cells is lo-cated in the nucleus, where it has a role in regulatingthe availability of EIF4E for the cytoplasmic translationalmachinery, by retaining EIF4E in the nucleus [16]. Highnuclear levels of 4EBP1 would thus inhibit translationand subsequent proliferation, which may explain its rela-tion with a good prognosis. The associations betweencytoplasmic 4EBP1 as well as high mRNA levels withhigh grade and poor prognosis indicate a dual role forthis protein. 4EBP1 has recently been implicated in apositive feedback loop by binding and stabilising mTORC1,thereby promoting its activation [12]. In the present study,p4EBP1 expression was correlated with pAKT_S473 butnot with pmTOR_S2448, a site associated with mTORC1[42]. Furthermore, recent studies have indicated additionalroles of 4EBP1, independent of mTORC1. Rapalogs, mainlytargeting mTORC1, have been shown to completely inhibitpS6K but only to partially inhibit p4EBP1 [2]. In bladdercancer, 4EBP1 was shown to be regulated by PI3K but notthrough mTORC1 [43], and mTOR-independent 4EBP1phosphorylation has been associated with resistance tomTOR kinase inhibitors [44]. Additional kinases for 4EBP1regulation remain to be identified. Upstream factors of thePI3K/AKT pathway are likely candidates. Some studieshave shown that mTOR kinase inhibitors block p4EBP1more effectively than rapalogs [2], suggesting mTORC2 asa candidate in 4EBP1 regulation. In our material, there is asignificant correlation between cytoplasmic p21-activatedkinase 1 (PAK1) and p4EBP1 (data not shown) and thearea around S65 in 4EBP1 is in agreement with the


consensus sequence reported for PAK1 [45], addingPAK1 to the list of potential candidates. Interestingly,PAK1 was recently described as involved in mTORC2mediated AKT_S473 phosphorylation, and the kinasemay be a part of the complex [46].Upregulation of the PI3K/AKT/mTOR pathway has

been associated with decreased benefit from endocrinetherapies in breast cancer, and recent studies supportmTOR inhibitors as promising agents for overcomingendocrine resistance [4]. Also, nuclear S6K2 has been as-sociated with response to endocrine therapy, althoughdependent on PgR status [8]. In our present study, highcytoplasmic but not nuclear expression of 4EBP1 predictedless benefit from tamoxifen, which reached significance for4EBP1 but not for p4EBP1. 4EBP1 is regulated by phos-phorylation at multiple sites, and the role for the differentsites is not totally established. The 4EBP1 antibody used inour study is raised towards a sequence surrounding S112,thus at the very C-terminus of 4EBP1, and recognises bothunphosphorylated as well as 4EBP1 phosphorylated at dif-ferent sites. In addition, the 4EBP1 and p4EBP1_S65 stain-ings are highly correlated, especially for the cytoplasmicpools of the proteins (Additional file 6), indicating that tosome extent the same proteins are detected. This may alsoreflect that an increase in total protein expression is oftenaccompanied with an increased phosphorylation and activa-tion of the proteins. 4EBP1 activation (foremost by mTOR)may therefore be the reason behind its role in endocrine re-sistance. Interestingly, in a recently published study, bothphosphorylated and total 4EBP1 were related to a poor out-come among patients with ER-positive breast cancers,treated with tamoxifen [47], in keeping with our findings.In that study, protein expression was determined by reverse-phase protein arrays, ruling out the possibility to distinguishbetween cytoplasmic and nuclear expression.In the present study, the predictive value for 4EBP1

was especially evident in the ER/PgR-expressing sub-group. In addition, the prognostic significance of 4EBP1was most prominent in combination with PgR expres-sion, suggesting a possible cross-talk between 4EBP1and nuclear receptors. The role of progesterone signal-ling in breast cancer remains controversial. In general,circulating progesterone is considered a risk factor forbreast cancer development by promoting cellular prolif-eration. However, in primary breast cancer, PgR expres-sion is associated with differentiated, less aggressivetumours and a favourable prognosis [48]. Upregulationof the insulin-like growth factor/PI3K/AKT/mTOR path-way is one suggested mechanism behind PgR downregula-tion in breast cancer, despite a functional ER. In agreement,our study showed an inverse association between S6K2/4EBP1 and PgR mRNA levels, in the three available co-horts. Furthermore, the gene encoding PgR is located atthe proximal part of the 11q chromosomal arm, which is

commonly deleted in 11q13/8p12 amplified tumours [11].However, 4EBP1 was recently described as a possible tar-get gene for PgR [49], suggesting the presence of a nega-tive feedback loop downregulating PgR after growth factorpathway stimulation. The function of PgR can be regulatedby receptor phosphorylation at multiple sites, throughgrowth factor receptor signalling pathways, and a subpop-ulation of cytoplasmic PgR has also been shown able toactivate kinase cascades, including PI3K/AKT [48]. It istempting to speculate that a coordinated expression ofPgR and cytoplasmic growth signalling factors includingS6K2/4EBP1 may facilitate the proliferative and oncogenicrole of PgR, promoting tumour progression and therapyresistance. In addition, PgR may in the long run be down-regulated through PI3K/AKT/mTOR pathway stimulationand subsequent aberrant ER signalling, leading to acquiredendocrine resistance among patients with initially ER/PgR-positive breast cancers.

ConclusionsInhibitors of mTOR signalling may have a clinical potentialin the management of several malignancies, not least as acomplement to ER-targeted therapies in breast cancer.However, the complexity of mTOR signalling is far fromunravelled. This study evaluates the clinical value of mTOReffectors in breast cancer. We show that 4EBP1 mRNA ex-pression is correlated with S6K2 mRNA and that highS6K2 and/or 4EBP1 is associated with a poor outcome, infour different cohorts of breast cancer. In addition, highcytoplasmic 4EBP1 protein levels predicted a poor prog-nosis and a decreased benefit from tamoxifen in a largerandomised cohort. In summary, suggested pathways of4EBP1 are illustrated in Additional file 1: Figure S7. Al-together, we propose the mTOR effectors 4EBP1 andS6K2 as new potential clinical markers in breast cancer.

Additional files

Additional file 1: Figure S1. Showing patient flow through the study:the randomised Stockholm tamoxifen trial, Stockholm 2 and Stockholm 3cohorts. Tam, tamoxifen; RT, radiotherapy; CMF, cyclophosphamide–metotrexate–5-fluorouracil chemotherapy; TMA, tissue microarray; IHC,immunohistochemistry. Figure S2 showing examples of tumours gradedfor 4EBP1 nuclear and cytoplasmic staining: negative/weak (a); intermediate(b); and strong staining (c); and validation of 4EBP1 antibody specificityusing immunoblot with MCF7 cell lysate (d). Figure S3 showing examplesof tumours graded for p4EBP1_S65 nuclear and cytoplasmic staining:negative/weak (a); intermediate (b) and strong staining (c); and validationof p4EBP1_S65 antibody specificity; immunoblot using MCF7 cell lysate(d); p4EBP1 breast tumour tissue staining: control without lambda-phosphatase(e) and with lambda-phosphatase (f). Figure S4 showing Kaplan–Meiercurves and multivariate Cox regression of breast cancer survival (BCS) anddistant recurrence-free survival (DRFS) in the van de Vijver patient cohort, inrelation to: S6K1 mRNA (a); S6K2 mRNA; (b) S6K2 mRNA median (c); and4EBP1 mRNA (d). The Cox analysis included the following variables:adjuvant chemotherapy treatment, endocrine treatment, lymph node status,and ER status. Figure S5 showing Kaplan–Meier curves and multivariateCox regression of breast cancer survival (BCS) in the Karolinska patient

http://www.biomedcentral.com/content/supplementary/bcr3557-S1.ppt


cohort, in relation to: S6K1 mRNA (a); S6K2 mRNA (b); and 4EBP1mRNA (c).The Cox analysis included the following variables: adjuvant chemotherapytreatment, endocrine treatment, lymph node status, tumour size and ERstatus. Figure S6 showing Kaplan–Meier curves and multivariate Coxregression of breast cancer survival (BCS) in the Uppsala patient cohort,in relation to S6K1 mRNA (a); S6K2 mRNA (b); and 4EBP1mRNA (c). TheCox analysis included the following variables: adjuvant chemotherapytreatment, endocrine treatment, lymph node status, tumour size, and ERstatus. Figure S7 showing an overview of suggested 4EBP1 signallingpathways, based on results from this study and previous literature.

Additional file 2: Is Table S1 presenting an overview of the numberof patients in the different cohorts and samples available for thedifferent analyses. TMA, tissue microarray.

Additional file 3: Is Table S2 presenting patient characteristics ofthe different cohorts included in the study.

Additional file 4: Is Table S3 presenting the distribution of 4EBP1and p4EBP1_S65 protein expression among samples in theStockholm 3 cohort.

Additional file 5: Is Table S4 presenting 4EBP1, S6K1 and S6K2mRNA in the Stockholm 2 cohort: correlations to gene copynumber, clinicopathological factors, and the PI3K/AKT/mTORpathway. Tam, tamoxifen; RT, radiotherapy; CMF, cyclophosphamide–metotrexate–5-fluorouracil chemotherapy.

Additional file 6: Is Table S5 presenting 4EBP1 and p4EBP1_S65protein in the Stockholm 3 cohort: correlations toclinicopathological factors and the AKT/mTOR pathway.

AbbreviationsAKT: Protein kinase B/v-akt murine thymoma viral oncogene homologue;CCND1: Cyclin D1; 4EBP1: 4E-binding protein 1; ER: Oestrogen receptor alpha;ERBB2: v-erb-b2 erythroblastic leukemia viral oncogene homolog 2;HER2: Human epidermal growth factor receptor 2;IHC: Immunohistochemistry; mTOR: Mammalian/mechanistic target ofrapamycin; PAK1: p21-activated kinase 1; PCR: Polymerase chain reaction;p4EBP1: Phosphorylated 4EBP1; PgR: Progesterone receptor;PI3K: Phosphatidylinositol-3-kinase; S6K: p70 ribosomal S6 kinase.

Competing interestsThe authors declare that they have no competing interests.

Authors’ contributionsEK conceived of the study and was involved in the study design, carried outthe quantitative PCR analyses, participated in the IHC grading, performed thestatistical analyses including handling of the public datasets, participated ininterpretation of the results and drafted the manuscript. GP-T conceived ofthe study and was involved in the design of the study, carried out the IHCstainings and participated in interpretation of the results. RA was involved inthe design of the study, carried out the IHC stainings and participated ininterpretation of the results. JB and A-LH participated in the design of thestudy and participated in interpretation of the results. BN, LS and TF wereresponsible for the clinical trial, the collection of tumour tissue samples andIHC stainings of steroid hormone and HER2 receptors. DCS performed thehistological grading of the tumours from the Stockholm 3 cohort. OSconceived of the study, was responsible for the design and coordination ofthe study, participated in the statistical analyses and the result interpretation.All authors were involved in reading and revising the draft and approved thefinal manuscript.

AcknowledgementsThe authors thank Birgit Olsson, Birgitta Holmlund, Marie Ahnström-Waltersson, Linda Bojmar and Sander Ellegård for excellent technical helpwith tissue preparation, Jelle ten Hoeve and Yudi Pawitan for contributingwith information regarding the van de Vijver, Uppsala and Karolinska cohorts,and Helena Fohlin, Mika Gustafsson, Fredrik Barrenäs and Colm Nestor forvaluable guidance in the work with the public datasets.

Author details1Department of Clinical and Experimental Medicine, Division of Oncology,Linköping University, County Council of Östergötland, Linköping SE-58185,

Sweden. 2Department of Pathology and Cytology, Karolinska UniversityHospital, Solna, Stockholm SE-17176, Sweden. 3Department of Oncology,Karolinska University Hospital, Stockholm South General Hospital, StockholmSE-11883, Sweden. 4Department of Pathology, Molecular Pathology ResearchUnit, Massachusetts General Hospital, Boston, MA 02129, USA.

Received: 21 January 2013 Accepted: 25 September 2013Published: 17 October 2013

References1. Musgrove EA, Sutherland RL: Biological determinants of endocrine

resistance in breast cancer. Nat Rev Cancer 2009, 9:631–643.2. Laplante M, Sabatini DM: mTOR signaling in growth control and disease.

Cell 2012, 149:274–293.3. Miller TW, Balko JM, Arteaga CL: Phosphatidylinositol 3-kinase and

antiestrogen resistance in breast cancer. J Clin Oncol 2011, 29:4452–4461.4. Villarreal-Garza C, Cortes J, Andre F, Verma S: mTOR inhibitors in the

management of hormone receptor-positive breast cancer: the latestevidence and future directions. Ann Oncol 2012, 23:2526–2535.

5. Baselga J, Campone M, Piccart M, Burris HA 3rd, Rugo HS, Sahmoud T,Noguchi S, Gnant M, Pritchard KI, Lebrun F, Beck JT, Ito Y, Yardley D, Deleu I,Perez A, Bachelot T, Vittori L, Xu Z, Mukhopadhyay P, Lebwohl D, Hortobagyi GN:Everolimus in postmenopausal hormone-receptor-positive advanced breastcancer. N Engl J Med 2012, 366:520–529.

6. Lyzogubov V, Khozhaenko Y, Usenko V, Antonjuk S, Ovcharenko G,Tikhonkova I, Filonenko V: Immunohistochemical analysis of Ki-67, PCNAand S6K1/2 expression in human breast cancer. Exp Oncol 2005,27:141–144.

7. Couch FJ, Wang XY, Wu GJ, Qian J, Jenkins RB, James CD: Localization ofPS6K to chromosomal region 17q23 and determination of itsamplification in breast cancer. Cancer Res 1999, 59:1408–1411.

8. Perez-Tenorio G, Karlsson E, Waltersson MA, Olsson B, Holmlund B,Nordenskjöld B, Fornander T, Skoog L, Stål O: Clinical potential of themTOR targets S6K1 and S6K2 in breast cancer. Breast Cancer Res Treat2011, 128:713–723.

9. van de Vijver MJ, He YD, van't Veer LJ, Dai H, Hart AA, Voskuil DW, Schreiber GJ,Peterse JL, Roberts C, Marton MJ, Parrish M, Atsma D, Witteveen A, Glas A,Delahaye L, van der Velde T, Bartelink H, Rodenhuis S, Rutgers ET, Friend SH,Bernards R: A gene-expression signature as a predictor of survival in breastcancer. N Engl J Med 2002, 347:1999–2009.

10. Armengol G, Rojo F, Castellvi J, Iglesias C, Cuatrecasas M, Pons B, Baselga J:Ramon y Cajal S: 4E-binding protein 1: a key molecular ‘funnel factor’ inhuman cancer with clinical implications. Cancer Res 2007, 67:7551–7555.

11. Karlsson E, Waltersson MA, Bostner J, Perez-Tenorio G, Olsson B, Hallbeck AL,Stål O: High-resolution genomic analysis of the 11q13 amplicon in breastcancers identifies synergy with 8p12 amplification, involving the mTORtargets S6K2 and 4EBP1. Genes Chromosomes Cancer 2011, 50:775–787.

12. Wang L, Rhodes CJ, Lawrence JC Jr: Activation of mammalian target ofrapamycin (mTOR) by insulin is associated with stimulation of 4EBP1binding to dimeric mTOR complex 1. J Biol Chem 2006, 281:24293–24303.

13. Ma XM, Yoon SO, Richardson CJ, Julich K, Blenis J: SKAR links pre-mRNAsplicing to mTOR/S6K1-mediated enhanced translation efficiency ofspliced mRNAs. Cell 2008, 133:303–313.

14. Bachmann RA, Kim JH, Wu AL, Park IH, Chen J: A nuclear transport signalin mammalian target of rapamycin is critical for its cytoplasmic signalingto S6 kinase 1. J Biol Chem 2006, 281:7357–7363.

15. Goh ET, Pardo OE, Michael N, Niewiarowski A, Totty N, Volkova D, TsanevaIR, Seckl MJ, Gout I: Involvement of heterogeneous ribonucleoprotein F inthe regulation of cell proliferation via the mammalian target ofrapamycin/S6 kinase 2 pathway. J Biol Chem 2010, 285:17065–17076.

16. Rong L, Livingstone M, Sukarieh R, Petroulakis E, Gingras AC, Crosby K,Smith B, Polakiewicz RD, Pelletier J, Ferraiuolo MA, Sonenburg N: Control ofeIF4E cellular localization by eIF4E-binding proteins, 4E-BPs. RNA 2008,14:1318–1327.

17. Yamnik RL, Digilova A, Davis DC, Brodt ZN, Murphy CJ, Holz MK: S6 kinase 1regulates estrogen receptor alpha in control of breast cancer cellproliferation. J Biol Chem 2009, 284:6361–6369.

18. McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM:Statistics Subcommittee of NCIEWGoCD: REporting recommendations fortumor MARKer prognostic studies (REMARK). Breast Cancer Res Treat 2006,100:229–235.

http://www.biomedcentral.com/content/supplementary/bcr3557-S2.pdf






19. Rutqvist LE, Johansson H: Long-term follow-up of the Stockholmrandomized trials of postoperative radiation therapy versus adjuvantchemotherapy among ‘high risk’ pre- and postmenopausal breast cancerpatients. Acta Oncol 2006, 45:517–527.

20. Rutqvist LE, Johansson H: Stockholm Breast Cancer Study G: Long-termfollow-up of the randomized Stockholm trial on adjuvant tamoxifenamong postmenopausal patients with early stage breast cancer.Acta Oncol 2007, 46:133–145.

21. Gomez-Abad C, Pisonero H, Blanco-Aparicio C, Roncador G, Gonzalez-Menchen A,Martinez-Climent JA, Mata E, Rodriguez ME, Munoz-Gonzalez G, Sanchez-Beato M,Leal JF, Bischoff JR, Piris MA: PIM2 inhibition as a rational therapeuticapproach in B-cell lymphoma. Blood 2011, 118:5517–5527.

22. Jerevall PL, Ma XJ, Li H, Salunga R, Kesty NC, Erlander MG, Sgroi DC, Holmlund B,Skoog L, Fornander T, Nordenskjöld B, Stål O: Prognostic utility of HOXB13:IL17BR and molecular grade index in early-stage breast cancer patientsfrom the Stockholm trial. Br J Cancer 2011, 104:1762–1769.

23. Khoshnoud MR, Löfdahl B, Fohlin H, Fornander T, Stål O, Skoog L, Bergh J,Nordenskjöld B: Immunohistochemistry compared to cytosol assays fordetermination of estrogen receptor and prediction of the long-termeffect of adjuvant tamoxifen. Breast Cancer Res Treat 2011, 126:421–430.

24. Stål O, Sullivan S, Sun XF, Wingren S, Nordenskjöld B: Simultaneousanalysis of c-erbB-2 expression and DNA content in breast cancer usingflow cytometry. Cytometry 1994, 16:160–168.

25. Gunnarsson C, Ahnström M, Kirschner K, Olsson B, Nordenskjöld B, Rutqvist LE,Skoog L, Stål O: Amplification of HSD17B1 and ERBB2 in primary breastcancer. Oncogene 2003, 22:34–40.

26. Jansson A, Delander L, Gunnarsson C, Fornander T, Skoog L, Nordenskjöld B,Stål O: Ratio of 17HSD1 to 17HSD2 protein expression predicts theoutcome of tamoxifen treatment in postmenopausal breast cancerpatients. Clin Cancer Res 2009, 15:3610–3616.

27. Stål O, Sullivan S, Wingren S, Skoog L, Rutqvist LE, Carstensen JM:Nordenskjöld B: c-erbB-2 expression and benefit from adjuvantchemotherapy and radiotherapy of breast cancer. Eur J Cancer 1995,31A:2185–2190.

28. Perez-Tenorio G, Alkhori L, Olsson B, Waltersson MA, Nordenskjöld B,Rutqvist LE, Skoog L, Stål O: PIK3CA mutations and PTEN loss correlatewith similar prognostic factors and are not mutually exclusive in breastcancer. Clin Cancer Res 2007, 13:3577–3584.

29. Stål O, Perez-Tenorio G, Åkerberg L, Olsson B, Nordenskjöld B, Skoog L,Rutqvist LE: Akt kinases in breast cancer and the results of adjuvanttherapy. Breast Cancer Res 2003, 5:R37–R44.

30. Bostner J, Karlsson E, Pandiyan MJ, Westman H, Skoog L, Fornander T,Nordenskjold B, Stal O: Activation of Akt, mTOR, and the estrogenreceptor as a signature to predict tamoxifen treatment benefit. BreastCancer Res Treat 2012, 137:397–406.

31. http://bioinformatics.nki.nl/data.php.32. Miller LD, Smeds J, George J, Vega VB, Vergara L, Ploner A, Pawitan Y, Hall P,

Klaar S, Liu ET, Bergh J: An expression signature for p53 status in humanbreast cancer predicts mutation status, transcriptional effects, andpatient survival. Proc Natl Acad Sci U S A 2005, 102:13550–13555.

33. Pawitan Y, Bjohle J, Amler L, Borg AL, Egyhazi S, Hall P, Han X, Holmberg L,Huang F, Klaar S, Liu ET, Miller L, Nordgren H, Ploner A, Sandelin K, ShawPM, Smeds J, Skoog L, Wedrén S, Bergh J: Gene expression profilingspares early breast cancer patients from adjuvant therapy: derived andvalidated in two population-based cohorts. Breast Cancer Res 2005,7:R953–R964.

34. Filonenko VV, Tytarenko R, Azatjan SK, Savinska LO, Gaydar YA, Gout IT,Usenko VS, Lyzogubov VV: Immunohistochemical analysis of S6K1 andS6K2 localization in human breast tumors. Exp Oncol 2004, 26:294–299.

35. Maruani DM, Spiegel TN, Harris EN, Shachter AS, Unger HA, Herrero-Gonzalez S, Holz MK: Estrogenic regulation of S6K1 expression creates apositive regulatory loop in control of breast cancer cell proliferation.Oncogene 2012, 31:5073–5080.

36. Benavente S, Verges R, Hermosilla E, Fumanal V, Casanova N, Garcia A,Ramon YCS, Giralt J: Overexpression of phosphorylated 4E-BP1 predictsfor tumor recurrence and reduced survival in cervical carcinoma treatedwith postoperative radiotherapy. Int J Radiat Oncol Biol Phys 2009,75:1316–1322.

37. Castellvi J, Garcia A, Rojo F, Ruiz-Marcellan C, Gil A, Baselga J: Ramon y Cajal S:Phosphorylated 4E binding protein 1: a hallmark of cell signaling thatcorrelates with survival in ovarian cancer. Cancer 2006, 107:1801–1811.

38. O'Reilly KE, Warycha M, Davies MA, Rodrik V, Zhou XK, Yee H, Polsky D,Pavlick AC, Rosen N, Bhardwaj N, Mills G, Osman I: Phosphorylated 4E-BP1is associated with poor survival in melanoma. Clin Cancer Res 2009,15:2872–2878.

39. Rojo F, Najera L, Lirola J, Jimenez J, Guzman M, Sabadell MD, Baselga J:Ramon y Cajal S: 4E-binding protein 1, a cell signaling hallmark in breastcancer that correlates with pathologic grade and prognosis. Clin CancerRes 2007, 13:81–89.

40. Balakumaran BS, Porrello A, Hsu DS, Glover W, Foye A, Leung JY, Sullivan BA,Hahn WC, Loda M, Febbo PG: MYC activity mitigates response to rapamycinin prostate cancer through eukaryotic initiation factor 4E-binding protein1-mediated inhibition of autophagy. Cancer Res 2009, 69:7803–7810.

41. Vogel C, Marcotte EM: Insights into the regulation of protein abundancefrom proteomic and transcriptomic analyses. Nat Rev Genet 2012,13:227–232.

42. Copp J, Manning G, Hunter T: TORC-specific phosphorylation ofmammalian target of rapamycin (mTOR): phospho-Ser2481 is a markerfor intact mTOR signaling complex 2. Cancer Res 2009, 69:1821–1827.

43. Nawroth R, Stellwagen F, Schulz WA, Stoehr R, Hartmann A, Krause BJ,Gschwend JE, Retz M: S6K1 and 4E-BP1 are independent regulated andcontrol cellular growth in bladder cancer. PLoS One 2011, 6:e27509.

44. Zhang Y, Zheng XF: mTOR-independent 4E-BP1 phosphorylation isassociated with cancer resistance to mTOR kinase inhibitors. Cell Cycle2012, 11:594–603.

45. Rennefahrt UE, Deacon SW, Parker SA, Devarajan K, Beeser A, Chernoff J,Knapp S, Turk BE, Peterson JR: Specificity profiling of Pak kinases allowsidentification of novel phosphorylation sites. J Biol Chem 2007,282:15667–15678.

46. Riaz A, Zeller KS, Johansson S: Receptor-specific mechanisms regulatephosphorylation of AKT at Ser473: role of RICTOR in β1 integrin-mediatedcell survival. PLoS One 2012, 7:e32081.

47. Meric-Bernstam F, Chen H, Akcakanat A, Do KA, Lluch A, Hennessy BT,Hortobagyi GN, Mills GB, Gonzalez-Angulo AM: Aberrations in translationalregulation are associated with poor prognosis in hormone receptor-positivebreast cancer. Breast Cancer Res 2012, 14:R138.

48. Obr AE, Edwards DP: The biology of progesterone receptor in the normalmammary gland and in breast cancer. Mol Cell Endocrinol 2012, 357:4–17.

49. Wood CE, Register TC, Cline JM: Transcriptional profiles of progestogeneffects in the postmenopausal breast. Breast Cancer Res Treat 2009,114:233–242.

doi:10.1186/bcr3557Cite this article as: Karlsson et al.: The mTOR effectors 4EBP1 and S6K2are frequently coexpressed, and associated with a poor prognosis andendocrine resistance in breast cancer: a retrospective study includingpatients from the randomised Stockholm tamoxifen trials. Breast CancerResearch 2013 15:R96.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

http://bioinformatics.nki.nl/data.php

the mtor effectors 4ebp1 and s6k2 are frequently...

Documents